{
  "best_metric": 1.403451919555664,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.02408405334617816,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00012042026673089081,
      "grad_norm": 51.93952178955078,
      "learning_rate": 1e-05,
      "loss": 8.9258,
      "step": 1
    },
    {
      "epoch": 0.00012042026673089081,
      "eval_loss": 2.869473695755005,
      "eval_runtime": 1613.5833,
      "eval_samples_per_second": 8.668,
      "eval_steps_per_second": 2.167,
      "step": 1
    },
    {
      "epoch": 0.00024084053346178162,
      "grad_norm": 60.6890869140625,
      "learning_rate": 2e-05,
      "loss": 9.6361,
      "step": 2
    },
    {
      "epoch": 0.00036126080019267245,
      "grad_norm": 49.853965759277344,
      "learning_rate": 3e-05,
      "loss": 9.3639,
      "step": 3
    },
    {
      "epoch": 0.00048168106692356323,
      "grad_norm": 44.821109771728516,
      "learning_rate": 4e-05,
      "loss": 9.3872,
      "step": 4
    },
    {
      "epoch": 0.0006021013336544541,
      "grad_norm": 52.650264739990234,
      "learning_rate": 5e-05,
      "loss": 8.6714,
      "step": 5
    },
    {
      "epoch": 0.0007225216003853449,
      "grad_norm": 48.42646408081055,
      "learning_rate": 6e-05,
      "loss": 8.5036,
      "step": 6
    },
    {
      "epoch": 0.0008429418671162356,
      "grad_norm": 41.25947952270508,
      "learning_rate": 7e-05,
      "loss": 7.4928,
      "step": 7
    },
    {
      "epoch": 0.0009633621338471265,
      "grad_norm": 35.24884033203125,
      "learning_rate": 8e-05,
      "loss": 7.9738,
      "step": 8
    },
    {
      "epoch": 0.0010837824005780173,
      "grad_norm": 56.005191802978516,
      "learning_rate": 9e-05,
      "loss": 7.6931,
      "step": 9
    },
    {
      "epoch": 0.0012042026673089081,
      "grad_norm": 55.710784912109375,
      "learning_rate": 0.0001,
      "loss": 8.2487,
      "step": 10
    },
    {
      "epoch": 0.001324622934039799,
      "grad_norm": 30.49041748046875,
      "learning_rate": 9.999316524962345e-05,
      "loss": 7.5405,
      "step": 11
    },
    {
      "epoch": 0.0014450432007706898,
      "grad_norm": 27.38810157775879,
      "learning_rate": 9.997266286704631e-05,
      "loss": 7.4251,
      "step": 12
    },
    {
      "epoch": 0.0015654634675015804,
      "grad_norm": 33.649085998535156,
      "learning_rate": 9.993849845741524e-05,
      "loss": 7.4058,
      "step": 13
    },
    {
      "epoch": 0.0016858837342324713,
      "grad_norm": 25.100475311279297,
      "learning_rate": 9.989068136093873e-05,
      "loss": 7.6971,
      "step": 14
    },
    {
      "epoch": 0.001806304000963362,
      "grad_norm": 29.93914031982422,
      "learning_rate": 9.98292246503335e-05,
      "loss": 5.9399,
      "step": 15
    },
    {
      "epoch": 0.001926724267694253,
      "grad_norm": 46.186336517333984,
      "learning_rate": 9.975414512725057e-05,
      "loss": 6.8558,
      "step": 16
    },
    {
      "epoch": 0.0020471445344251438,
      "grad_norm": 25.15791130065918,
      "learning_rate": 9.966546331768191e-05,
      "loss": 6.7501,
      "step": 17
    },
    {
      "epoch": 0.0021675648011560346,
      "grad_norm": 31.10689353942871,
      "learning_rate": 9.956320346634876e-05,
      "loss": 7.1959,
      "step": 18
    },
    {
      "epoch": 0.0022879850678869254,
      "grad_norm": 28.72784423828125,
      "learning_rate": 9.944739353007344e-05,
      "loss": 6.7444,
      "step": 19
    },
    {
      "epoch": 0.0024084053346178163,
      "grad_norm": 30.710111618041992,
      "learning_rate": 9.931806517013612e-05,
      "loss": 7.9429,
      "step": 20
    },
    {
      "epoch": 0.002528825601348707,
      "grad_norm": 25.784887313842773,
      "learning_rate": 9.917525374361912e-05,
      "loss": 7.3364,
      "step": 21
    },
    {
      "epoch": 0.002649245868079598,
      "grad_norm": 28.78515625,
      "learning_rate": 9.901899829374047e-05,
      "loss": 6.82,
      "step": 22
    },
    {
      "epoch": 0.0027696661348104888,
      "grad_norm": 39.985713958740234,
      "learning_rate": 9.884934153917997e-05,
      "loss": 7.9684,
      "step": 23
    },
    {
      "epoch": 0.0028900864015413796,
      "grad_norm": 28.817001342773438,
      "learning_rate": 9.86663298624003e-05,
      "loss": 7.268,
      "step": 24
    },
    {
      "epoch": 0.00301050666827227,
      "grad_norm": 29.924957275390625,
      "learning_rate": 9.847001329696653e-05,
      "loss": 7.3812,
      "step": 25
    },
    {
      "epoch": 0.003130926935003161,
      "grad_norm": 25.788951873779297,
      "learning_rate": 9.826044551386744e-05,
      "loss": 7.6471,
      "step": 26
    },
    {
      "epoch": 0.0032513472017340517,
      "grad_norm": 41.178375244140625,
      "learning_rate": 9.803768380684242e-05,
      "loss": 7.2271,
      "step": 27
    },
    {
      "epoch": 0.0033717674684649425,
      "grad_norm": 44.785343170166016,
      "learning_rate": 9.780178907671789e-05,
      "loss": 8.1423,
      "step": 28
    },
    {
      "epoch": 0.0034921877351958334,
      "grad_norm": 61.3057861328125,
      "learning_rate": 9.755282581475769e-05,
      "loss": 7.6013,
      "step": 29
    },
    {
      "epoch": 0.003612608001926724,
      "grad_norm": 38.59844970703125,
      "learning_rate": 9.729086208503174e-05,
      "loss": 7.5294,
      "step": 30
    },
    {
      "epoch": 0.003733028268657615,
      "grad_norm": 39.695777893066406,
      "learning_rate": 9.701596950580806e-05,
      "loss": 6.7086,
      "step": 31
    },
    {
      "epoch": 0.003853448535388506,
      "grad_norm": 21.75993537902832,
      "learning_rate": 9.672822322997305e-05,
      "loss": 6.1454,
      "step": 32
    },
    {
      "epoch": 0.003973868802119396,
      "grad_norm": 23.870563507080078,
      "learning_rate": 9.642770192448536e-05,
      "loss": 5.9876,
      "step": 33
    },
    {
      "epoch": 0.0040942890688502875,
      "grad_norm": 20.87793731689453,
      "learning_rate": 9.611448774886924e-05,
      "loss": 5.3824,
      "step": 34
    },
    {
      "epoch": 0.004214709335581178,
      "grad_norm": 18.55045509338379,
      "learning_rate": 9.578866633275288e-05,
      "loss": 4.9466,
      "step": 35
    },
    {
      "epoch": 0.004335129602312069,
      "grad_norm": 19.357398986816406,
      "learning_rate": 9.545032675245813e-05,
      "loss": 5.4976,
      "step": 36
    },
    {
      "epoch": 0.00445554986904296,
      "grad_norm": 19.688756942749023,
      "learning_rate": 9.509956150664796e-05,
      "loss": 5.8604,
      "step": 37
    },
    {
      "epoch": 0.004575970135773851,
      "grad_norm": 27.641071319580078,
      "learning_rate": 9.473646649103818e-05,
      "loss": 5.2527,
      "step": 38
    },
    {
      "epoch": 0.004696390402504741,
      "grad_norm": 25.47074317932129,
      "learning_rate": 9.43611409721806e-05,
      "loss": 5.385,
      "step": 39
    },
    {
      "epoch": 0.0048168106692356326,
      "grad_norm": 20.537351608276367,
      "learning_rate": 9.397368756032445e-05,
      "loss": 5.119,
      "step": 40
    },
    {
      "epoch": 0.004937230935966523,
      "grad_norm": 21.654861450195312,
      "learning_rate": 9.357421218136386e-05,
      "loss": 4.6382,
      "step": 41
    },
    {
      "epoch": 0.005057651202697414,
      "grad_norm": 22.191814422607422,
      "learning_rate": 9.316282404787871e-05,
      "loss": 5.1202,
      "step": 42
    },
    {
      "epoch": 0.005178071469428305,
      "grad_norm": 19.029094696044922,
      "learning_rate": 9.273963562927695e-05,
      "loss": 5.2586,
      "step": 43
    },
    {
      "epoch": 0.005298491736159196,
      "grad_norm": 18.765365600585938,
      "learning_rate": 9.230476262104677e-05,
      "loss": 4.8687,
      "step": 44
    },
    {
      "epoch": 0.005418912002890086,
      "grad_norm": 17.537826538085938,
      "learning_rate": 9.185832391312644e-05,
      "loss": 4.6707,
      "step": 45
    },
    {
      "epoch": 0.0055393322696209776,
      "grad_norm": 24.50990867614746,
      "learning_rate": 9.140044155740101e-05,
      "loss": 5.1278,
      "step": 46
    },
    {
      "epoch": 0.005659752536351868,
      "grad_norm": 25.02142906188965,
      "learning_rate": 9.093124073433463e-05,
      "loss": 5.2466,
      "step": 47
    },
    {
      "epoch": 0.005780172803082759,
      "grad_norm": 22.66439437866211,
      "learning_rate": 9.045084971874738e-05,
      "loss": 5.0619,
      "step": 48
    },
    {
      "epoch": 0.00590059306981365,
      "grad_norm": 22.0762939453125,
      "learning_rate": 8.995939984474624e-05,
      "loss": 5.0657,
      "step": 49
    },
    {
      "epoch": 0.00602101333654454,
      "grad_norm": 25.61806297302246,
      "learning_rate": 8.945702546981969e-05,
      "loss": 4.1101,
      "step": 50
    },
    {
      "epoch": 0.00602101333654454,
      "eval_loss": 1.7742722034454346,
      "eval_runtime": 1624.1807,
      "eval_samples_per_second": 8.612,
      "eval_steps_per_second": 2.153,
      "step": 50
    },
    {
      "epoch": 0.006141433603275431,
      "grad_norm": 47.616703033447266,
      "learning_rate": 8.894386393810563e-05,
      "loss": 6.9972,
      "step": 51
    },
    {
      "epoch": 0.006261853870006322,
      "grad_norm": 27.753664016723633,
      "learning_rate": 8.842005554284296e-05,
      "loss": 6.9823,
      "step": 52
    },
    {
      "epoch": 0.006382274136737213,
      "grad_norm": 25.53302574157715,
      "learning_rate": 8.788574348801675e-05,
      "loss": 6.5077,
      "step": 53
    },
    {
      "epoch": 0.006502694403468103,
      "grad_norm": 29.319658279418945,
      "learning_rate": 8.73410738492077e-05,
      "loss": 6.7296,
      "step": 54
    },
    {
      "epoch": 0.006623114670198995,
      "grad_norm": 25.24061393737793,
      "learning_rate": 8.678619553365659e-05,
      "loss": 5.6474,
      "step": 55
    },
    {
      "epoch": 0.006743534936929885,
      "grad_norm": 21.104772567749023,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.9952,
      "step": 56
    },
    {
      "epoch": 0.006863955203660776,
      "grad_norm": 20.799846649169922,
      "learning_rate": 8.564642241456986e-05,
      "loss": 6.0229,
      "step": 57
    },
    {
      "epoch": 0.006984375470391667,
      "grad_norm": 15.64926528930664,
      "learning_rate": 8.506183921362443e-05,
      "loss": 5.9788,
      "step": 58
    },
    {
      "epoch": 0.007104795737122558,
      "grad_norm": 19.641931533813477,
      "learning_rate": 8.44676704559283e-05,
      "loss": 6.0515,
      "step": 59
    },
    {
      "epoch": 0.007225216003853448,
      "grad_norm": 23.41942024230957,
      "learning_rate": 8.386407858128706e-05,
      "loss": 6.7682,
      "step": 60
    },
    {
      "epoch": 0.00734563627058434,
      "grad_norm": 17.75417709350586,
      "learning_rate": 8.32512286056924e-05,
      "loss": 6.5946,
      "step": 61
    },
    {
      "epoch": 0.00746605653731523,
      "grad_norm": 20.17172622680664,
      "learning_rate": 8.262928807620843e-05,
      "loss": 6.3011,
      "step": 62
    },
    {
      "epoch": 0.007586476804046121,
      "grad_norm": 18.58847999572754,
      "learning_rate": 8.199842702516583e-05,
      "loss": 6.2966,
      "step": 63
    },
    {
      "epoch": 0.007706897070777012,
      "grad_norm": 17.919443130493164,
      "learning_rate": 8.135881792367686e-05,
      "loss": 5.8059,
      "step": 64
    },
    {
      "epoch": 0.007827317337507902,
      "grad_norm": 19.31273651123047,
      "learning_rate": 8.07106356344834e-05,
      "loss": 6.4417,
      "step": 65
    },
    {
      "epoch": 0.007947737604238793,
      "grad_norm": 25.046363830566406,
      "learning_rate": 8.005405736415126e-05,
      "loss": 6.5747,
      "step": 66
    },
    {
      "epoch": 0.008068157870969685,
      "grad_norm": 23.319303512573242,
      "learning_rate": 7.938926261462366e-05,
      "loss": 6.6744,
      "step": 67
    },
    {
      "epoch": 0.008188578137700575,
      "grad_norm": 22.34912872314453,
      "learning_rate": 7.871643313414718e-05,
      "loss": 6.3949,
      "step": 68
    },
    {
      "epoch": 0.008308998404431465,
      "grad_norm": 39.79279327392578,
      "learning_rate": 7.803575286758364e-05,
      "loss": 6.6698,
      "step": 69
    },
    {
      "epoch": 0.008429418671162356,
      "grad_norm": 20.666879653930664,
      "learning_rate": 7.734740790612136e-05,
      "loss": 6.3727,
      "step": 70
    },
    {
      "epoch": 0.008549838937893248,
      "grad_norm": 22.23097038269043,
      "learning_rate": 7.66515864363997e-05,
      "loss": 6.7035,
      "step": 71
    },
    {
      "epoch": 0.008670259204624138,
      "grad_norm": 20.92945098876953,
      "learning_rate": 7.594847868906076e-05,
      "loss": 6.7143,
      "step": 72
    },
    {
      "epoch": 0.008790679471355029,
      "grad_norm": 22.52288055419922,
      "learning_rate": 7.52382768867422e-05,
      "loss": 7.2588,
      "step": 73
    },
    {
      "epoch": 0.00891109973808592,
      "grad_norm": 26.24117660522461,
      "learning_rate": 7.452117519152542e-05,
      "loss": 7.2383,
      "step": 74
    },
    {
      "epoch": 0.009031520004816811,
      "grad_norm": 23.615114212036133,
      "learning_rate": 7.379736965185368e-05,
      "loss": 7.0135,
      "step": 75
    },
    {
      "epoch": 0.009151940271547702,
      "grad_norm": 31.925701141357422,
      "learning_rate": 7.30670581489344e-05,
      "loss": 7.4073,
      "step": 76
    },
    {
      "epoch": 0.009272360538278592,
      "grad_norm": 27.146015167236328,
      "learning_rate": 7.233044034264034e-05,
      "loss": 6.8409,
      "step": 77
    },
    {
      "epoch": 0.009392780805009483,
      "grad_norm": 27.483409881591797,
      "learning_rate": 7.158771761692464e-05,
      "loss": 8.3212,
      "step": 78
    },
    {
      "epoch": 0.009513201071740375,
      "grad_norm": 28.326284408569336,
      "learning_rate": 7.083909302476453e-05,
      "loss": 7.1594,
      "step": 79
    },
    {
      "epoch": 0.009633621338471265,
      "grad_norm": 31.824556350708008,
      "learning_rate": 7.008477123264848e-05,
      "loss": 5.7829,
      "step": 80
    },
    {
      "epoch": 0.009754041605202156,
      "grad_norm": 19.395477294921875,
      "learning_rate": 6.932495846462261e-05,
      "loss": 6.1882,
      "step": 81
    },
    {
      "epoch": 0.009874461871933046,
      "grad_norm": 21.194475173950195,
      "learning_rate": 6.855986244591104e-05,
      "loss": 5.6866,
      "step": 82
    },
    {
      "epoch": 0.009994882138663936,
      "grad_norm": 29.51766014099121,
      "learning_rate": 6.778969234612584e-05,
      "loss": 6.1513,
      "step": 83
    },
    {
      "epoch": 0.010115302405394828,
      "grad_norm": 16.740205764770508,
      "learning_rate": 6.701465872208216e-05,
      "loss": 4.9738,
      "step": 84
    },
    {
      "epoch": 0.010235722672125719,
      "grad_norm": 14.368413925170898,
      "learning_rate": 6.623497346023418e-05,
      "loss": 4.8067,
      "step": 85
    },
    {
      "epoch": 0.01035614293885661,
      "grad_norm": 14.684015274047852,
      "learning_rate": 6.545084971874738e-05,
      "loss": 4.3561,
      "step": 86
    },
    {
      "epoch": 0.0104765632055875,
      "grad_norm": 15.932342529296875,
      "learning_rate": 6.466250186922325e-05,
      "loss": 4.7409,
      "step": 87
    },
    {
      "epoch": 0.010596983472318392,
      "grad_norm": 16.245309829711914,
      "learning_rate": 6.387014543809223e-05,
      "loss": 4.7687,
      "step": 88
    },
    {
      "epoch": 0.010717403739049282,
      "grad_norm": 16.515871047973633,
      "learning_rate": 6.307399704769099e-05,
      "loss": 4.7443,
      "step": 89
    },
    {
      "epoch": 0.010837824005780173,
      "grad_norm": 41.284793853759766,
      "learning_rate": 6.227427435703997e-05,
      "loss": 4.6509,
      "step": 90
    },
    {
      "epoch": 0.010958244272511063,
      "grad_norm": 19.552005767822266,
      "learning_rate": 6.147119600233758e-05,
      "loss": 4.4044,
      "step": 91
    },
    {
      "epoch": 0.011078664539241955,
      "grad_norm": 18.036300659179688,
      "learning_rate": 6.066498153718735e-05,
      "loss": 5.0476,
      "step": 92
    },
    {
      "epoch": 0.011199084805972846,
      "grad_norm": 17.481563568115234,
      "learning_rate": 5.985585137257401e-05,
      "loss": 4.7952,
      "step": 93
    },
    {
      "epoch": 0.011319505072703736,
      "grad_norm": 14.768298149108887,
      "learning_rate": 5.90440267166055e-05,
      "loss": 4.1727,
      "step": 94
    },
    {
      "epoch": 0.011439925339434626,
      "grad_norm": 18.626522064208984,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 5.0798,
      "step": 95
    },
    {
      "epoch": 0.011560345606165518,
      "grad_norm": 16.772817611694336,
      "learning_rate": 5.74131823855921e-05,
      "loss": 4.5098,
      "step": 96
    },
    {
      "epoch": 0.011680765872896409,
      "grad_norm": 17.050249099731445,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 4.3496,
      "step": 97
    },
    {
      "epoch": 0.0118011861396273,
      "grad_norm": 19.16489028930664,
      "learning_rate": 5.577423184847932e-05,
      "loss": 4.4216,
      "step": 98
    },
    {
      "epoch": 0.01192160640635819,
      "grad_norm": 20.456478118896484,
      "learning_rate": 5.495227651252315e-05,
      "loss": 4.1681,
      "step": 99
    },
    {
      "epoch": 0.01204202667308908,
      "grad_norm": 17.859432220458984,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 4.5742,
      "step": 100
    },
    {
      "epoch": 0.01204202667308908,
      "eval_loss": 1.48580801486969,
      "eval_runtime": 1625.6909,
      "eval_samples_per_second": 8.604,
      "eval_steps_per_second": 2.151,
      "step": 100
    },
    {
      "epoch": 0.012162446939819972,
      "grad_norm": 16.177490234375,
      "learning_rate": 5.330452921628497e-05,
      "loss": 5.4763,
      "step": 101
    },
    {
      "epoch": 0.012282867206550863,
      "grad_norm": 16.521148681640625,
      "learning_rate": 5.247918773366112e-05,
      "loss": 5.9816,
      "step": 102
    },
    {
      "epoch": 0.012403287473281753,
      "grad_norm": 14.81554126739502,
      "learning_rate": 5.165316846586541e-05,
      "loss": 5.6084,
      "step": 103
    },
    {
      "epoch": 0.012523707740012643,
      "grad_norm": 13.736647605895996,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 5.2882,
      "step": 104
    },
    {
      "epoch": 0.012644128006743536,
      "grad_norm": 18.304468154907227,
      "learning_rate": 5e-05,
      "loss": 5.8812,
      "step": 105
    },
    {
      "epoch": 0.012764548273474426,
      "grad_norm": 18.765226364135742,
      "learning_rate": 4.917330276168208e-05,
      "loss": 5.8377,
      "step": 106
    },
    {
      "epoch": 0.012884968540205316,
      "grad_norm": 15.608325958251953,
      "learning_rate": 4.834683153413459e-05,
      "loss": 5.9864,
      "step": 107
    },
    {
      "epoch": 0.013005388806936207,
      "grad_norm": 14.807425498962402,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 5.8354,
      "step": 108
    },
    {
      "epoch": 0.013125809073667099,
      "grad_norm": 17.752962112426758,
      "learning_rate": 4.669547078371504e-05,
      "loss": 6.3213,
      "step": 109
    },
    {
      "epoch": 0.01324622934039799,
      "grad_norm": 16.0859432220459,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 6.3305,
      "step": 110
    },
    {
      "epoch": 0.01336664960712888,
      "grad_norm": 17.609752655029297,
      "learning_rate": 4.504772348747687e-05,
      "loss": 6.2805,
      "step": 111
    },
    {
      "epoch": 0.01348706987385977,
      "grad_norm": 18.305526733398438,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 6.1058,
      "step": 112
    },
    {
      "epoch": 0.013607490140590662,
      "grad_norm": 15.66598129272461,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 6.2276,
      "step": 113
    },
    {
      "epoch": 0.013727910407321553,
      "grad_norm": 18.535301208496094,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 6.6646,
      "step": 114
    },
    {
      "epoch": 0.013848330674052443,
      "grad_norm": 20.5335693359375,
      "learning_rate": 4.17702704859633e-05,
      "loss": 6.6563,
      "step": 115
    },
    {
      "epoch": 0.013968750940783333,
      "grad_norm": 18.792814254760742,
      "learning_rate": 4.095597328339452e-05,
      "loss": 6.7214,
      "step": 116
    },
    {
      "epoch": 0.014089171207514224,
      "grad_norm": 18.318065643310547,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 6.6231,
      "step": 117
    },
    {
      "epoch": 0.014209591474245116,
      "grad_norm": 15.907029151916504,
      "learning_rate": 3.933501846281267e-05,
      "loss": 6.3682,
      "step": 118
    },
    {
      "epoch": 0.014330011740976006,
      "grad_norm": 17.899784088134766,
      "learning_rate": 3.852880399766243e-05,
      "loss": 6.7449,
      "step": 119
    },
    {
      "epoch": 0.014450432007706897,
      "grad_norm": 20.180458068847656,
      "learning_rate": 3.772572564296005e-05,
      "loss": 7.5801,
      "step": 120
    },
    {
      "epoch": 0.014570852274437787,
      "grad_norm": 30.31849479675293,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 7.2028,
      "step": 121
    },
    {
      "epoch": 0.01469127254116868,
      "grad_norm": 21.378841400146484,
      "learning_rate": 3.612985456190778e-05,
      "loss": 7.0255,
      "step": 122
    },
    {
      "epoch": 0.01481169280789957,
      "grad_norm": 18.160831451416016,
      "learning_rate": 3.533749813077677e-05,
      "loss": 7.235,
      "step": 123
    },
    {
      "epoch": 0.01493211307463046,
      "grad_norm": 19.299564361572266,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 6.5939,
      "step": 124
    },
    {
      "epoch": 0.01505253334136135,
      "grad_norm": 22.814464569091797,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 7.6817,
      "step": 125
    },
    {
      "epoch": 0.015172953608092243,
      "grad_norm": 19.72553253173828,
      "learning_rate": 3.298534127791785e-05,
      "loss": 7.2951,
      "step": 126
    },
    {
      "epoch": 0.015293373874823133,
      "grad_norm": 27.33616828918457,
      "learning_rate": 3.221030765387417e-05,
      "loss": 8.9711,
      "step": 127
    },
    {
      "epoch": 0.015413794141554023,
      "grad_norm": 17.395479202270508,
      "learning_rate": 3.144013755408895e-05,
      "loss": 5.0004,
      "step": 128
    },
    {
      "epoch": 0.015534214408284914,
      "grad_norm": 23.039718627929688,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 5.8222,
      "step": 129
    },
    {
      "epoch": 0.015654634675015804,
      "grad_norm": 18.903642654418945,
      "learning_rate": 2.991522876735154e-05,
      "loss": 5.7666,
      "step": 130
    },
    {
      "epoch": 0.015775054941746695,
      "grad_norm": 17.474811553955078,
      "learning_rate": 2.916090697523549e-05,
      "loss": 5.5946,
      "step": 131
    },
    {
      "epoch": 0.015895475208477585,
      "grad_norm": 15.69929027557373,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 5.485,
      "step": 132
    },
    {
      "epoch": 0.01601589547520848,
      "grad_norm": 14.750075340270996,
      "learning_rate": 2.766955965735968e-05,
      "loss": 5.3102,
      "step": 133
    },
    {
      "epoch": 0.01613631574193937,
      "grad_norm": 14.924285888671875,
      "learning_rate": 2.693294185106562e-05,
      "loss": 5.0332,
      "step": 134
    },
    {
      "epoch": 0.01625673600867026,
      "grad_norm": 18.159025192260742,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 5.2681,
      "step": 135
    },
    {
      "epoch": 0.01637715627540115,
      "grad_norm": 17.145938873291016,
      "learning_rate": 2.547882480847461e-05,
      "loss": 4.6567,
      "step": 136
    },
    {
      "epoch": 0.01649757654213204,
      "grad_norm": 18.082134246826172,
      "learning_rate": 2.476172311325783e-05,
      "loss": 4.8821,
      "step": 137
    },
    {
      "epoch": 0.01661799680886293,
      "grad_norm": 16.63163185119629,
      "learning_rate": 2.405152131093926e-05,
      "loss": 5.1819,
      "step": 138
    },
    {
      "epoch": 0.01673841707559382,
      "grad_norm": 13.199299812316895,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 4.443,
      "step": 139
    },
    {
      "epoch": 0.016858837342324712,
      "grad_norm": 12.945577621459961,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 4.4436,
      "step": 140
    },
    {
      "epoch": 0.016979257609055606,
      "grad_norm": 14.055652618408203,
      "learning_rate": 2.196424713241637e-05,
      "loss": 4.9036,
      "step": 141
    },
    {
      "epoch": 0.017099677875786496,
      "grad_norm": 14.148738861083984,
      "learning_rate": 2.128356686585282e-05,
      "loss": 4.2584,
      "step": 142
    },
    {
      "epoch": 0.017220098142517386,
      "grad_norm": 15.948631286621094,
      "learning_rate": 2.061073738537635e-05,
      "loss": 4.9077,
      "step": 143
    },
    {
      "epoch": 0.017340518409248277,
      "grad_norm": 14.8971586227417,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 5.1003,
      "step": 144
    },
    {
      "epoch": 0.017460938675979167,
      "grad_norm": 17.40970802307129,
      "learning_rate": 1.928936436551661e-05,
      "loss": 4.8937,
      "step": 145
    },
    {
      "epoch": 0.017581358942710058,
      "grad_norm": 14.42086410522461,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 4.09,
      "step": 146
    },
    {
      "epoch": 0.017701779209440948,
      "grad_norm": 13.88598346710205,
      "learning_rate": 1.800157297483417e-05,
      "loss": 4.2703,
      "step": 147
    },
    {
      "epoch": 0.01782219947617184,
      "grad_norm": 17.1777286529541,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 4.753,
      "step": 148
    },
    {
      "epoch": 0.01794261974290273,
      "grad_norm": 22.961992263793945,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 6.3804,
      "step": 149
    },
    {
      "epoch": 0.018063040009633623,
      "grad_norm": 20.709712982177734,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 5.3321,
      "step": 150
    },
    {
      "epoch": 0.018063040009633623,
      "eval_loss": 1.4118815660476685,
      "eval_runtime": 1624.7275,
      "eval_samples_per_second": 8.609,
      "eval_steps_per_second": 2.152,
      "step": 150
    },
    {
      "epoch": 0.018183460276364513,
      "grad_norm": 15.155057907104492,
      "learning_rate": 1.553232954407171e-05,
      "loss": 5.3383,
      "step": 151
    },
    {
      "epoch": 0.018303880543095404,
      "grad_norm": 16.10227394104004,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 5.5482,
      "step": 152
    },
    {
      "epoch": 0.018424300809826294,
      "grad_norm": 14.932662010192871,
      "learning_rate": 1.435357758543015e-05,
      "loss": 5.9801,
      "step": 153
    },
    {
      "epoch": 0.018544721076557184,
      "grad_norm": 14.543388366699219,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 6.0454,
      "step": 154
    },
    {
      "epoch": 0.018665141343288075,
      "grad_norm": 13.57608413696289,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 5.3455,
      "step": 155
    },
    {
      "epoch": 0.018785561610018965,
      "grad_norm": 14.566596984863281,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 5.8565,
      "step": 156
    },
    {
      "epoch": 0.018905981876749856,
      "grad_norm": 16.764638900756836,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 6.2423,
      "step": 157
    },
    {
      "epoch": 0.01902640214348075,
      "grad_norm": 13.570207595825195,
      "learning_rate": 1.157994445715706e-05,
      "loss": 5.244,
      "step": 158
    },
    {
      "epoch": 0.01914682241021164,
      "grad_norm": 15.006532669067383,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 5.8121,
      "step": 159
    },
    {
      "epoch": 0.01926724267694253,
      "grad_norm": 16.283781051635742,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 5.7995,
      "step": 160
    },
    {
      "epoch": 0.01938766294367342,
      "grad_norm": 16.97859764099121,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 6.0864,
      "step": 161
    },
    {
      "epoch": 0.01950808321040431,
      "grad_norm": 16.730764389038086,
      "learning_rate": 9.549150281252633e-06,
      "loss": 6.1446,
      "step": 162
    },
    {
      "epoch": 0.0196285034771352,
      "grad_norm": 21.55950164794922,
      "learning_rate": 9.068759265665384e-06,
      "loss": 6.9383,
      "step": 163
    },
    {
      "epoch": 0.019748923743866092,
      "grad_norm": 15.633070945739746,
      "learning_rate": 8.599558442598998e-06,
      "loss": 5.7383,
      "step": 164
    },
    {
      "epoch": 0.019869344010596982,
      "grad_norm": 15.562016487121582,
      "learning_rate": 8.141676086873572e-06,
      "loss": 6.0859,
      "step": 165
    },
    {
      "epoch": 0.019989764277327873,
      "grad_norm": 16.343961715698242,
      "learning_rate": 7.695237378953223e-06,
      "loss": 6.5659,
      "step": 166
    },
    {
      "epoch": 0.020110184544058766,
      "grad_norm": 15.342779159545898,
      "learning_rate": 7.260364370723044e-06,
      "loss": 5.9446,
      "step": 167
    },
    {
      "epoch": 0.020230604810789657,
      "grad_norm": 17.29735565185547,
      "learning_rate": 6.837175952121306e-06,
      "loss": 6.8397,
      "step": 168
    },
    {
      "epoch": 0.020351025077520547,
      "grad_norm": 18.98801612854004,
      "learning_rate": 6.425787818636131e-06,
      "loss": 7.2175,
      "step": 169
    },
    {
      "epoch": 0.020471445344251438,
      "grad_norm": 18.11264419555664,
      "learning_rate": 6.026312439675552e-06,
      "loss": 7.3273,
      "step": 170
    },
    {
      "epoch": 0.020591865610982328,
      "grad_norm": 17.66663932800293,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 5.952,
      "step": 171
    },
    {
      "epoch": 0.02071228587771322,
      "grad_norm": 16.35909652709961,
      "learning_rate": 5.263533508961827e-06,
      "loss": 6.7025,
      "step": 172
    },
    {
      "epoch": 0.02083270614444411,
      "grad_norm": 19.418554306030273,
      "learning_rate": 4.900438493352055e-06,
      "loss": 7.5946,
      "step": 173
    },
    {
      "epoch": 0.020953126411175,
      "grad_norm": 17.824628829956055,
      "learning_rate": 4.549673247541875e-06,
      "loss": 6.5499,
      "step": 174
    },
    {
      "epoch": 0.021073546677905893,
      "grad_norm": 18.88226318359375,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 7.4214,
      "step": 175
    },
    {
      "epoch": 0.021193966944636784,
      "grad_norm": 21.682384490966797,
      "learning_rate": 3.885512251130763e-06,
      "loss": 7.1637,
      "step": 176
    },
    {
      "epoch": 0.021314387211367674,
      "grad_norm": 23.528362274169922,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 8.0835,
      "step": 177
    },
    {
      "epoch": 0.021434807478098564,
      "grad_norm": 29.83986473083496,
      "learning_rate": 3.271776770026963e-06,
      "loss": 9.481,
      "step": 178
    },
    {
      "epoch": 0.021555227744829455,
      "grad_norm": 19.354703903198242,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 5.6199,
      "step": 179
    },
    {
      "epoch": 0.021675648011560345,
      "grad_norm": 13.174699783325195,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 4.2994,
      "step": 180
    },
    {
      "epoch": 0.021796068278291236,
      "grad_norm": 22.209993362426758,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 5.7409,
      "step": 181
    },
    {
      "epoch": 0.021916488545022126,
      "grad_norm": 13.499473571777344,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 5.0113,
      "step": 182
    },
    {
      "epoch": 0.022036908811753016,
      "grad_norm": 16.739299774169922,
      "learning_rate": 1.962316193157593e-06,
      "loss": 5.207,
      "step": 183
    },
    {
      "epoch": 0.02215732907848391,
      "grad_norm": 14.593423843383789,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 5.5588,
      "step": 184
    },
    {
      "epoch": 0.0222777493452148,
      "grad_norm": 18.718555450439453,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 5.8553,
      "step": 185
    },
    {
      "epoch": 0.02239816961194569,
      "grad_norm": 14.348808288574219,
      "learning_rate": 1.333670137599713e-06,
      "loss": 5.3079,
      "step": 186
    },
    {
      "epoch": 0.02251858987867658,
      "grad_norm": 13.264379501342773,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 4.1346,
      "step": 187
    },
    {
      "epoch": 0.022639010145407472,
      "grad_norm": 14.749774932861328,
      "learning_rate": 9.810017062595322e-07,
      "loss": 4.5342,
      "step": 188
    },
    {
      "epoch": 0.022759430412138362,
      "grad_norm": 13.615583419799805,
      "learning_rate": 8.247462563808817e-07,
      "loss": 4.5825,
      "step": 189
    },
    {
      "epoch": 0.022879850678869253,
      "grad_norm": 14.845542907714844,
      "learning_rate": 6.819348298638839e-07,
      "loss": 4.9798,
      "step": 190
    },
    {
      "epoch": 0.023000270945600143,
      "grad_norm": 13.39441204071045,
      "learning_rate": 5.526064699265753e-07,
      "loss": 4.9815,
      "step": 191
    },
    {
      "epoch": 0.023120691212331037,
      "grad_norm": 15.792435646057129,
      "learning_rate": 4.367965336512403e-07,
      "loss": 4.9431,
      "step": 192
    },
    {
      "epoch": 0.023241111479061927,
      "grad_norm": 11.874963760375977,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 4.2139,
      "step": 193
    },
    {
      "epoch": 0.023361531745792818,
      "grad_norm": 13.765231132507324,
      "learning_rate": 2.458548727494292e-07,
      "loss": 4.555,
      "step": 194
    },
    {
      "epoch": 0.023481952012523708,
      "grad_norm": 12.91434383392334,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 4.3116,
      "step": 195
    },
    {
      "epoch": 0.0236023722792546,
      "grad_norm": 15.871813774108887,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 4.6468,
      "step": 196
    },
    {
      "epoch": 0.02372279254598549,
      "grad_norm": 14.559955596923828,
      "learning_rate": 6.150154258476315e-08,
      "loss": 4.6366,
      "step": 197
    },
    {
      "epoch": 0.02384321281271638,
      "grad_norm": 14.423089981079102,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 4.4505,
      "step": 198
    },
    {
      "epoch": 0.02396363307944727,
      "grad_norm": 16.981185913085938,
      "learning_rate": 6.834750376549792e-09,
      "loss": 4.2474,
      "step": 199
    },
    {
      "epoch": 0.02408405334617816,
      "grad_norm": 15.380423545837402,
      "learning_rate": 0.0,
      "loss": 4.0541,
      "step": 200
    },
    {
      "epoch": 0.02408405334617816,
      "eval_loss": 1.403451919555664,
      "eval_runtime": 1622.9561,
      "eval_samples_per_second": 8.618,
      "eval_steps_per_second": 2.155,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.475451310300201e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}